T+1 Across EU, UK and Switzerland: One Coordinated Cutover, Three Rulebooks

The three markets picked the same date. They did not pick the same rulebook — and a single group has to satisfy all three from one cutover weekend.

The convenient story is that Europe moves to T+1 on 11 October 2027, so you treat it as one change and staff one programme. The story is convenient because it lets you reuse the North American playbook and copy the migration you already ran for the May 2024 move to T+1 in the United States. It is also wrong in a way that surfaces late — in the second half of the programme, when the assumption of a single regime is baked into your target operating model and expensive to unpick. The EU, the United Kingdom and Switzerland aligned on the date. They did not align on the instrument that mandates it, the discipline regime that penalises failure, or the body that will ask you to prove you were ready. A group active in all three has to run one coordinated cutover against three separate rulebooks, and the interesting engineering is in the reconciliation.

One date, three legal instruments

The EU move is statutory. Regulation (EU) 2025/2075, which amends the Central Securities Depositories Regulation, was published in the Official Journal in October 2025 and sets the standard settlement cycle for in-scope transactions at T+1 through an amended Article 5. It is hard law, with Level 2 settlement-discipline measures behind it, and ESMA published its final report on the amended settlement-discipline technical standards in October 2025.

The UK reaches the same date by a different route. HM Treasury has legislated the mandatory cutover through an amendment to the onshored UK CSDR, but the operational substance sits in the UK T+1 Code of Conduct produced under the Accelerated Settlement Taskforce — a set of recommended practices, not a directly enforceable rulebook. The date is binding; much of the behaviour around it is guidance.

Switzerland has no equivalent single securities regulator issuing the rule. The move is coordinated through industry self-regulation — the Swiss Securities Post-Trade Council and the exchange rulebooks — explicitly mirroring the EU and UK timetable so that the Swiss market does not desynchronise from the venues its participants also trade. Same destination, three constitutional starting points: a regulation, a statutory instrument dressed with a code, and a set of self-regulatory recommendations. That difference in legal texture is not academic. It changes who you have to satisfy, what counts as evidence, and what happens when you fail.

Where the rulebooks actually diverge

The point that matters operationally is the penalty for failing to settle, because that is what your exception process is optimising against. Here the three regimes are genuinely different.

The EU carries a statutory cash-penalty regime under CSDR settlement discipline. Fails accrue daily penalties, calculated and collected through the CSDs, with the rate structure being recalibrated as part of the T+1 package and the Commission retaining the ability to adjust or suspend the mechanism if it threatens market stability. A fail in an EU CSD is a measurable, invoiced cost, and your economics have to model it. The UK does not replicate that statutory cash-penalty machinery; its discipline leans on the code of conduct and existing market-infrastructure charges rather than a CSDR-style penalty regime imported wholesale. Switzerland has signalled no intention to introduce a cash-penalty regime for fails at all.

The consequence is concrete. The same operational fail — an unmatched instruction, a missing allocation, a stale standing instruction — has a hard financial cost in the EU, a softer reputational-and-charges cost in the UK, and effectively no direct penalty in Switzerland. If you build one exception queue and prioritise every fail identically, you either over-invest in chasing Swiss fails that cost nothing or under-invest in EU fails that are billing you daily. The penalty asymmetry has to be a field in your data model, not a footnote in the policy. We have written separately on instrumenting the exception queue for CSDR penalties; the cross-border version of that problem is the same queue carrying three different cost functions.

Scope: mostly aligned, with edges that bite

On scope the three are close, which is precisely what makes the gaps dangerous — they are small enough to be missed. All three move on-venue equities, exchange-traded products and bonds that currently settle T+2 onto T+1. Securities financing transactions are carved out in the EU and the UK on comparable terms, and UK gilts already settle T+1, so they are not part of the change there.

The edges are where an instrument is in scope in one regime and treated differently in another, or where an OTC leg tied to an on-venue trade is expected to follow the accelerated cycle by convention rather than by rule. The failure mode is assuming the carve-outs match line for line and discovering, in testing, that a product family moved in one jurisdiction and not in the neighbouring one. The correct assumption is convergence, not identity: build the scope determination per venue and per settlement location, and let the answer differ.

Design one programme, not three

The instinct on seeing three rulebooks is to run three workstreams. That is the second error, and it is as expensive as pretending the regimes are identical. The affected machinery — allocation and confirmation timing, standing settlement instructions, funding and FX, the fails process — is shared plumbing. Splitting it by jurisdiction duplicates the build and creates three subtly different versions of the same control, which then have to be reconciled anyway.

The workable shape is one programme with jurisdiction as a parameter. The compressed post-trade timeline is common to all three: allocations and confirmations have to be same-day, so the latency budget for allocations and confirmations is a single design regardless of which rulebook applies. Standing settlement instructions have to be clean before cutover in every market, so treat SSIs as master data once, for the whole estate. What varies — penalty economics, evidentiary expectations, the enforceability of the code versus the regulation — becomes configuration and reporting differences layered on that common base, not separate builds. The settlement-readiness programme for the middle office is the same programme in all three markets; the rulebook is a lookup, not a fork.

Governance follows the same logic. One steering group owns the cutover, one cutover runbook coordinates the weekend across the three markets, and one readiness dashboard reports status with a jurisdiction dimension. Three programmes give you three runbooks that have to agree with each other at the seams — which they will not, on the night.

The reconciliation you actually own

Reduce this to a decision and it is not “how do we comply with three rulebooks.” It is “where do we hold the differences.” The right answer is: in a small number of well-marked places — a scope table, a penalty-cost field, an evidence map — sitting on top of one shared operating model, so that the common plumbing is built once and the divergences are visible, tested and owned. The wrong answers are the two symmetrical failures: one programme that assumes the regimes are the same and is blindsided by the penalty asymmetry, or three programmes that duplicate the plumbing and never quite reconcile at the boundary. The markets did the coordination for you by agreeing the date. The reconciliation of the three rulebooks is the part they left on your desk.